In this paper, a novel deinterlacing algorithm is proposed. The proposed algorithm approximates the distribution of the luminance into a polynomial function. Instead of using one polynomial function for all pixels, different polynomial functions are used for the uniform, texture, and directional edge regions. The function coefficients for each region are computed by matrix multiplications. Experimental results demonstrate that the proposed method performs better than the conventional algorithms.
|Number of pages||3|
|Journal||World Academy of Science, Engineering and Technology|
|Publication status||Published - 2009 Jan 1|
- Polynomial interpolation
ASJC Scopus subject areas